*****read me*****
***to recreate paper results***
Three datasets are needed.

1. credit files available from one of three credit agencies. Contact author or editorial office for contact information to purchase data.
2. Census 2000 summary files. These are publically available information of population information by census block/tract/msa, etc. 
See http://factfinder.census.gov/servlet/DatasetMainPageServlet for information on files and instructions for purchase. 
3. Payday lending information and geocodes are available from the website of professor Steven Graves at Cal State Northridge.
http://www.csun.edu/~sg4002/research/usury.html

These come in various formats. Each requires different extraction methods. Once the credit file is available, export to a fixed delimit file and import using programs here.
Similarly, use the census extraction program (provided with purchase of data) to export the needed fields. Import using programs here. Finally, repeat with payday loan data.

Files to relabel variables are also available here.

Run the 'clean' program to ensure that variables coded with error fields are removed.

Run the popstats program to calculate distances.


***create figures****
figure 2 is simply the predicted values from table 3, column 2 for %black==1 and %white==1.

****create tables****
use file 'mainregs'
